mark kibort

running a car at near stoich at part throttle conditions and low power settings is not an issue. not enough heat to cause pinging, usually. its when the the power setting increases, thats when the mixture needs to change. one of the reasons the older cars have the fuel enrichment switch at 70% of fullthrottle . before that, the O2 sensor is keeping the fuel air ration hunting from full rich to full lean. when it goes open loop, you get the ratios near the 13:1 range. without the fuel enrichment switch, you can operate the engine at near stoich at WOT, and that can be bad. I did tests in that area. no pinging, but i wouldnt trust operating at those ratios for very long

Mk

docmirror

Brian, sorry, it's complicated subject, and made more so by long held ideas from before we were able to accurately measure what's going on inside a cylinder. The short answer from me is 'I don't know'. I would not presume to recommend any A/F ratio for a modified 928 engine being supercharged. I know what you want to achieve, but since you're a test engineer with this engine, all the settings have to be best guesses until you get some real world data to analyze.

The only thing I'm saying, and it's backed up very well by data from the aircraft engine 'tuners' is that the most dangerous area of operation is just slightly rich of peak EGT. Operations lean of peak EGT are more forgiving of overboost because the potential for detonation is much lower. However, if you look at the chart on post 1, you will see that there is a cost as well. You can't get max HP lean, because, quite simply there isn't enough fuel to burn to make HP. If you want to run rich of peak, it's fine, but do it about 100F rich, not 25F, which, surprisingly, doesn't cost HP(middle red line of BHP)!

Look at the top chart of the graph in post 1. Note that peak EGT occurs about 1525 on the left scale. Follow that red peak to the left until it reaches 1425, about three grids to the left. Follow that grid line down to the middle scale, and you will see that your BHP is just a bit higher when slightly over-rich, follow that grid line further down and note that the specific fuel consumption is a bit higher than the absolute minimum.

Where danger lies. The most dangerous range of operation is WOT, tuned for 25-75 deg F rich of peak. Looking at the graphs, it's also just about where max HP is produced. If you're gonna run rich, run quite rich under high MP.

I will make a bold guess and say that your initial settings for F/A sound spot on. I might go a bit richer for the make-power, around 11.5:1 but that's your call. The reason is you can alway lean out slightly if it reacts the way you want, but if you blow it to bits, you can't subsequently richen it, cause there will be motor parts all over the garage floor. May I suggest welding in a EGT bung on your exhaust, about 2-3" away from the flange, and also getting a CHT for the hottest cylinder? That way you can use your results on the dyno to both tune for optimum HP while staying out of the knock danger area.

Best of luck.

Doc

a4sfed928

Question if you are running lean of Peak and decide you want max power Don't you have to pass through Peak to get to the ROP condition to make max power? If this is the case for more transient operation you would be at peak more often than running at ROP so more heat.....Scotty you know the drill.

worf928

This is only true in your case since you don't have your system hooked to vacuum. For the rest of us, the pressure in the rails is effected by the vacuum. You can verify this by hooking a fuel pressure gauge to your rails after you reconnect your vacuum lines. The WSM and Spec books have specs for rail pressure with and without vacuum. Why are they different if the vacuum has no effect on fuel pressure? And the effect is not just an idle effect.

However, you did indirectly point out something I misstated. The vacuum on the fuel pressure isn't a 'load signal' it is more of a second derivative of the load. The vacuum drops precipitously when the throttle plate is opening. The recovery of vacuum to a steady non-zero state, if any, is proportional to the final steady-state load. Again, you can test this too.

You are missing my point. No you cannot prove 'this' (at least my point) on the dyno. On the dyno there is only one open-throttle-event. Other than for the first fraction of a second the throttle plate is wide open for the entire dyno run - unless your purpose is tuning part-throttle parameters.

Note that I have not disputed any posts about steady-state operation whether that operation is part-throttle or WOT. I am talking about the transition from one throttle state to another.

13.1 is a richer than stoich mixture. So, your point is?

That's because the control system knows what it is trying to do and can do a pretty good job of it despite the fact that you've disconnected one of its actuators.

I have never, ever, ever, seen a 928 dyno run with a stock LH system where the a/f ratio was constant for the entire run. Actually, no, that's not true. I did see one - the ratio was 14.7:1 for the whole run. Because the idle switch was broken and keeping the 02 loop on all the time.

I'm not saying that such a dyno chart doesn't exist. But, I've seen multiple-dozen, if not a hundred, and never seen a constant a/f ratio for a non-broken car.

There will be no vacuum until the throttle plate stops moving. And only if the plate stops moving before WOT. Try it. You'll see. If you start at idle with vacuum then open the throttle by a fraction (1/3 or 1/4) you will see vacuum drop and then recover shortly after the plate stops moving. If you keep the plate fully open (like on a dyno run) then there will be no vacuum recovery.

Please spend the extra minute to format your posts to make them readable. Yeah - sounds a little pissy I know. But, if you won't spend a minute making your posts readable why should I spend 5 minutes to decipher them? If we are trying to learn and/or reverse any misconceptions either of us may have clarity is important.

blown 87

If your F/A stays constant, under varying loads from idle to cruise, to WOT, you have a problem.

worf928

I don't know what charts your looking at Mark. The street 928 a/f charts from dyno runs (with non-broken cars) I've see have shown way richer than 12.6:1 a/f for stock set-ups. Low 12s and high 11s are common. If anything a lot of stock 928s could run a little leaner,make more power, and still not be leaner than 12.6:1.


The regulator does not 'provide' pressure. The pump provides the pressure. The regulator's function is to regulate pressure in the rails. This function is necessary since as load increases the volume of fuel leaving the rails via the injectors increases and as a result the pressure in the rails would otherwise drop. That's why the regulators and dampeners are hooked to vacuum. It's a simple way to offset what would otherwise be a pressure drop in the rails as load (and fuel usage) increases.

But, don't take my word for it. Go look at Probst or the Bosch Handbook.

This doesn't make any sense. Why would you need to pull back the pressure at idle? For one, at idle the lambda loop is in control and could deal with just about any steady pressure by changing injector pulse width. As for fuel economy, at idle your sitting still, or coasting, so what does that have to do with economy? If you meant emissions then I fall back to the 02 loop.

Ask the opposite question: why would the regulator increase fuel pressure off-idle. (Answer: it's a transient effect of the throttle opening.)

The way the regulator is intended to work is that the spring force is counter-balanced to some extent by the vacuum force. This allows the regulator to allow more or less fuel volume to return to the tank so as to (try to) keep pressure constant.

Maybe you need more fuel pressure from a RFPR because your normal FPR isn't able to regulate fuel pressure because you don't have the vacuum hooked up.

worf928

You may very well be right Doc. My knowledge stops at this point with respect to what happens in the chamber. At some point you decide to believe what's written in the text book etc. Sometimes it's wrong.

I agree. But, there's also tip-in detonation.

Let me ask you a question. Say you're flying in LOP. And you decide that you want to go faster. In fact, for whatever reason, you want to go as fast as possible. Do you just fire-wall the throttle?


Prior to the divergence to transient throttle response and a discussion of the 'physics' of the fuel pressure regulator one topic of this thread was the seemingly ubiquitous cautions to avoid overly-lean steady-state running conditions. One such caution was posted in this thread a few pages back. I've come across such cautions many times and until a few thread-pages back never really questioned those cautions. I don't know about anyone else, but I'd like to understand from where these cautions come. No where can I find an adequate explanation for these cautions.

The best explanations (and they were not much of an explanation) I've found for detonation of a lean mixture is due to the difficulty in getting overly-lean charges to mix well and combust well. And that was more in the context of transient response that steady-state operation.

worf928

You chose to respond to the description of the transient initial condition of my example and not the point of the example. My point is NOT that the control system cannot respond to the transient lean condition. It can. The point was ...

... to illustrate why the control system must respond (if responding as designed) to the increased load by going richer than stoich. If it richened the mixture only slightly or richened the mixture only to stoich then the potential for detonation would be much higher because of the decreased ignition energy requirement coupled with the sudden increase in charge density.

mark kibort

Hey, at off throttle, you are WAY lean of peak. you have to pass through it when you get on the gas, point is, it happens so quickly that the cylinder doesnt really even see stoich. you guys are forgeting that to fill a cylinder all the way, without the MAF responding for that air that has passed through its inlet, is almost impossible. again, its more likely that you partially fill a cylinder with less density vs the measured fuel flow for what was measured by the AFM or MAF. ever blip the throttle and get a stumble. sometimes this is a real rich condition where the fuel got there before the air. this could be complicated by the fact that LJet adn LHjet activate all the injectors at the same time, regardless of fireing order.

you guys are missing the point. the point is, is that there is no danger, as Doc says, of operation of lean of stoich, should your mixture go haywire or your NOS system forgets to turn on the gas. (unless this event causes a 14.7 ratio, for which it would be a Big problem, especially for long WOT durations)
during cruise, 14.7 is fine as there usually is not enough fuel to cause enough heat to hurt things, nor cause detonation. even less of an issue lean of stoich, where there is even less heat due to even less gas but still enough to give 80% of the power , but 10% less heat than during areas of richer operation for the same HP output.

again, we all race and drive with this situation and it is not an issue. full lean , to full rich and transient periods in between. this happens in almost all driving situations.

MK

mark kibort

dave,

please do the test as i have many times to prove my point.

Ive run it both ways, connected and disconnect. i put a vacuum guage on that vacuum source. guess what, when you give it just a little gas, the vacuum is gone. no vacuum, no effect at part throttle and beyond.

its not some closed loop system that uses vacume to control fuel pressure, again, that is done by the design of the spring loaded vacuum diaphram and fuel return line. the vacuum lines only reduce fuel pressure at idle and dont do anything during WOT or near WOT conditions.

an observation below:
1. if you saw a full throttle switch open, the fuel air ratios would still be hunting full rich to full lean
2. as far as formating the posts, it looked readable to me, with the darker and altered text. i havent learned how to do your fancy format to have the interactive discussion. sorry. teach me how with a PM

mk

mark kibort

none of this discussion, on my part, nor i believe on Doc's part, is saying to set your car to some lean of stoich operation. obviosly, to get the best power, you need to be rich of stoich. however, if your engine ever got caught at a way lean conditon, you wouldnt have an issue. real lean isnt a problem or danger for your engine . however, a problem that leans the mixture to near stoich, can be a big issue and could punch holes in the pistons.

mk

docmirror

Let's see if I can figure this out from a 928 perspective. Tip-in detonation is a very real possibility. That's why I didn't want to give advice about actual F/A ratios, because there are factors at work on pressurized 928 engines that I'm not too familiar with. As you correctly said before, you have to go through peak, to get to lean of peak(not really, you could run LOP all the time, but it's just not done in reality) where you can operate completely safe. It's that small segment of operation around rich of peak where trouble is waiting.

In a plane, when running LOP the pilot has absolute control over the manifold pressure(throttle) and F/A ratio with the mixture control, and the RPM with prop angle control. Most flying in complex aircraft is done with the throttle wide open above 5000' all the time, and the prop adjusted to the speed you want to go(higher RPM is faster). Most pilots get the setting they want, then adjust the mixture to some value slightly rich of peak. To accelerate, just speed up the prop, and the mixture will stay the same, if everything is working right. If you are LOP, it'll stay that way, if you are rich, it'll stay that way.

Now, the big question "avoid over-lean steady-state operation". This is a relative statement. when the optimum F/A ratio for a given power, and given safety margin is reached, it will be well rich of peak EGT. Notice that the engine produces it's max power when the F/A ratio is richer than peak EGT. Or, said another way, peak EGT is thermodynamically less efficient, which I think you understand quite well. So the statement 'over-lean' is a relative term meaning 'don't operate further on the upslope of rich of peak EGT'.

Having said that, once you go past peak, the temps and cylinder pressure both start down. This is all goodness, because it's more thermodynamically efficient, and it's also more chemically reactive efficient as you can see in the lowest graph in post 1, where BSFC is at it's lowest point, about one grid to the RIGHT of peak EGT.

So, in summary, when the discussion turns to 'overly-lean' don't equate that with Lean Of Peak EGT. Equate that statement with 'closer to peak EGT, in the lean direction, but still rich of peak, or possibly at peak EGT'. That's where engines explode, not running lean of peak EGT, but running leaner than max HP EGT. So referring back to the graph, the middle red line, just below the blue box, shows BHP, and it maxes out about three grid lines to the left of that magenta vertical line. Follow that grid line up from max HP, and you will see it intersects the EGT(top) graph at about 1425F EGT. A tuner would be very happy here. Now start leaning more and the HP stays fairly constant but the temps rise rapidly, and we quickly enter the area of likely detonation.

Hope this helps. I'll try to find another better graph and link to it later.

<edit: go to: http://www.gami.com/frames.htm and click on 'test results in the left column. This is an aircraft engine, but the results carry over to any spark-fired engine. You can also read the text: http://www.avweb.com/news/pelican/182544-1.html for a complete discussion of LOP operations as it relates to aircraft. It is very relevant for auto engines, but you have to wade through some pilot-speak to get the meat out>

Doc

mark kibort

GREAT statement. This is what most really are refering to when they talk about lean causing detonation. lean as the mixture approaches stoich! (or hits it or just past it) not way past it. Way past it, no harm can be found. Those charts are for most all engines, plane or car!

As Doc said, this is one of the most misunderstood concepts of auto engines. right up there with gear rations, air filters giving better mileage, and premium gas making more HP.

MK





doc said " So, in summary, when the discussion turns to 'overly-lean' don't equate that with Lean Of Peak EGT. Equate that statement with 'closer to peak EGT, in the lean direction, but still rich of peak, or possibly at peak EGT'. That's where engines explode, not running lean of peak EGT, but running leaner than max HP EGT. So referring back to the graph, the middle red line, just below the blue box, shows BHP, and it maxes out about three grid lines to the left of that magenta vertical line. Follow that grid line up from max HP, and you will see it intersects the EGT(top) graph at about 1425F EGT. A tuner would be very happy here. Now start leaning more and the HP stays fairly constant but the temps rise rapidly, and we quickly enter the area of likely detonation."

Larry928GTS

Measured it.

Binary responses?



Slightly longer, non-binary response for those that want it below:

Mark, every one of those statements that you made and I responded to are absolutely wrong. They're so wrong, and there are so many of them, that I have to think the statements were made as some kind of a troll attempt.

Fuel injector size is rated at a fairly standardized fuel pressure of 3 bar, which is 43.5psi. The injector flows the rated amount of fuel with that pressure differential across it. Changing the pressure differential across the injector will change the amount of fuel flow through it. Increasing the pressure differential across the injector will increase the amount of fuel flowing through it, and decreasing the pressure differential across it will reduce the amount of fuel flowing through it. The stock '87 and later 928 injectors are rated at 19lb/hr, but they don't flow 19lb/hr in a 928 engine. Why? Because the stock fuel pressure is higher than the 43.5psi that the injectors get rated at, so in a 928 engine they flow more than 19lb/hr.

Manifold vacuum is NOT an either all or nothing situation, as Mark stated. If it were, you'd have no part throttle, only idle and full throttle. Think about it. If the manifold pressure was the same just off idle as it is at full throttle, why would the car go faster at full throttle than it does just off idle? The pressure differential across the intake valves would be the same in either case, so airflow into the cylinders would be the same as well. I've seen quite a few vacuum gauges. They all had numbers on them, and not just "Yes" and "No" written on their faces for the needle to point to.

Since the fuel pressure is on one side of the injector, and the manifold pressure is on the other, fuel flow through the injector would vary as the manifold pressure changed, because the pressure differential across the injector would change. To help illustrate the point, think about what would happen if you had 50psi of fuel pressure in the fuel rail, and 60 psi of pressure in the manifold. When the injector opened, no fuel would flow through the injector. You'd instead have air blowing from the manifold side of the injector through it into the fuel rail.

The LH computer is mapped based on the fuel injectors having the same flow rating all the time, and not always bouncing up and down and all over the place. It's much easier and more accurate to have the LH computer control fuel flow into the engine when the injectors flow the same amount of fuel every time that they're opened for the same length of time. If not, the amount of fuel that would flow through the injector when it was opened for a specific time period could be different every time, even if the amount of time that the injector was opened was the same in each instance.

As already stated, manifold pressure changes with throttle position. Using just the spring inside of the fuel pressure to control the fuel pressure would result in the flow through the injector constantly changing as manifold pressure, and therefore the pressure differential across the injector, changed. The LH computer controls fuel delivery to the engine by varying the open time of the injectors. The LH is continuously changing the injector open time to change the fuel delivery to the engine based on input signals it's receiving and it's internal mapping. With the manifold pressure, and therefore pressure differential across the injector, constantly changing, using only the spring pressure of the fuel pressure regulator to control fuel pressure would be like you trying to adjust the water temperature in the shower while someone else was constantly messing with the hot water over in the next room.

The diaphram inside the fuel pressure regulator has a port on one side of it that connects to the manifold with a vacuum line. Vacuum in the manifold acts on the diaphram to counter some of the spring force, and reduce the fuel rail fuel pressure. As manifold pressure increases, the vacuum force on the diaphram decreases, allowing the spring to exert more force and increase the fuel pressure in the fuel rail. The stock diaphram mechanism is sized so as to provide a 1:1 ratio of change with the changes in manifold pressure that occur. This results in the fuel rail pressure changing in a 1:1 ratio with manifold pressure, and a constant pressure differential across the injector. Whenever the injector opening time is the same, the injector now flows the same amount of fuel through it every time. The flow through it doesn't go up and down with the manifold pressure for the same injector open time. The only thing controlling the fuel flow through the injector is the opening time as determined by the LH computer, and not the combination of the LH and pressure differential across the injector each doing their own thing.

Under idle and part throttle, the LH uses the injector opening times programmed into it's base maps. Based on feedback it receives from the O2 sensor, it "learns" and makes modifications to those base map opening times to achieve the desired 14.7:1 air/fuel ratio that gives lowest emissions with reasonable part throttle running and fuel economy. Under full throttle the LH ignores the O2 sensor signal. It uses injector opening times based on the learned part throttle opening times, along with the full throttle mapping offset added to that learned part throttle mapping. Change your fuel pressure with an adjustable fuel pressure regulator and the LH will counter the changes by using the O2 sensor feedback to modify the mapping and try to get back to the desired 14.7:1 air/fuel ratio. It will continue to do this with continued changes to the fuel pressure until the LH reaches it's limit of adjustability. Tune fuel pressure for maximum power at full throttle on the dyno, and the fuel delivery is based on the learned part throttle mapping offsets generated from driving before you got to the dyno, along with the full throttle mapping offset. Since the O2 sensor signal is ignored during the full throttle operation, no learning is occurring during this time to compensate for the changes being made to the fuel pressure for tuning purposes. After the full throttle tuning is completed, and the fuel pressure has been changed, you drive home under part throttle with that altered fuel pressure. The LH then begins "learning", and modifying the mapping to compensate for that changed fuel pressure. That now newly learned part throttle base map tune is then used to change the full throttle operation fuel delivery, changing it from what you had when you left the dyno. John Speake's SharkTuner manual states that it's important for the base fuel map to be correctly tuned before the full throttle mapping is done. There is a reason for it.

No need for him to even consult any of the books. There's been logged data from several different cars posted on this board a number of times that clearly show the relationship of manifold pressure on fuel pressure, and that manifold vacuum is not like some sort of on/off switch like Mark has stated. Mark, if your vacuum gauge really is showing zero vacuum at part throttle there are two possibilities. Your vacuum gauge is broken, or the the vacuum line is connected to the wrong location. My guess would be that you've got the vacuum line on the wrong port.

Four pages on this topic and nobody has even mentioned the effect of differing air/fuel ratio, charge temperature, and also pressure in the cylinder on combustion speed, and how differences in those factors change the combustion speed to affect the optimum ignition timing, knock, and pre-ignition?

BrianG

I was going to ask about the theory regarding the remapping of the EZK map with my SharkTuner, but you guys have so badly messed up my (mis)understanding of A/F issues that I'm thinking of just excising the whole ignition system and running without it!

That said, can we address some real world observations???

Given that we don't want to run just rich of stoich/peak EGT, under WOT (or even significant make-power throttle settings (>75%), it seems reasonable to map that portion to peak-power A/F of 12.6:1. How am I doing so far??

I have noticed that drivability suffers at part throttle (<75%) A/F mixtures much lean of stoich (perhaps past 15.5:1). I cannot see decent drivability at 17:1. Am I missing something?

If the above observation is valid, the discussion so far seem to indicate that mapping to stoich at these part throttle (<75%) settings will be safe. (There is no significant boost at these settings anyway.)

How does blood pressure, communicable disease, and intracranial space occupying lesion training qualify me to open the hood of this thing and pick up a wrench???